Difference between C89 and C99?

[Richard Bos]

23klines of diffs between the last draft and the released version?

No; between the last _public_ draft and the released version. AIUI there
has been another draft, possibly several, between n869 and the Standard,
but they were all FTCEO.

Richard

 

[those who know me have no need of my name]

in comp.lang.c i read:

Personally I want my code to be portable, so I'm not knowingly using any
extensions later than C89,

what about when your code is compiled by a c99 conforming compiler but
you've used something which does not work the same? fortunately the c99
rationale covers this -- i think completely, though i haven't exhaustively
checked -- and is free. e.g.,

,----[ page 41 ]
| unsigned long x;
| size_t y;
| x = y;
|
| silently truncates y.
`----{ ... if size_t is wider than unsigned long, nearly unheard of in c89 }


do you use nothing from amd1 either, e.g., useful wide-character support?

 

[Keith Thompson]

No; between the last _public_ draft and the released version. AIUI there
has been another draft, possibly several, between n869 and the Standard,
but they were all FTCEO.

FTCEO?

 

[Eric Sosman]

[...]

23klines of diffs between the last draft and the released version?

No; between the last _public_ draft and the released version. AIUI there
has been another draft, possibly several, between n869 and the Standard,
but they were all FTCEO.

FTCEO?

For The Conspirators' Eyes Only.

(Accuracy of transliteration not guaranteed -- it's
a QOI issue.)

 

[lawrence.jones]

23klines of diffs between the last draft and the released version?

No, between the last *public* draft and the released version. After
N869 there was N877 (a committee working document), the FDIS (Final
Draft International Standard, which is what was voted on by the powers
that be), and then the actual standard. And that includes header and
context lines, there are "only" 1042 hunks.

sounds like a recipe for releasing errors.

The vast majority of the changes were typesetting or indexing changes
(the index in N869 left a lot to be desired). That's what makes finding
the "interesting" changes challenging.

If they are like other diffs of text sources then most of them will be
meaningless differences in spacing and line wrapping.

Spacing and line wrapping have no significance in troff source files, so
there's little reason to change them unless the actual text changed.

But isn't there some version control which has a "changes since last
document" in English?

Not to that level of detail -- they tend to be things like "resolve
ballot comments" rather than a specific list of things that changed.
And they're per-file, they're not collated into an overall list. (For
what it's worth, the standard comprises 67 source files, one of which --
the index -- is automatically generated from the others.)

-Larry Jones

It's hard to be religious when certain people are never
incinerated by bolts of lightning. -- Calvin

 

[Richard Bos]

For The Conspirators' Eyes Only.

Unexpected accuracy.

Richard

 

[Richard Bos]

No, between the last *public* draft and the released version. After
N869 there was N877 (a committee working document), the FDIS (Final
Draft International Standard, which is what was voted on by the powers
that be), and then the actual standard. And that includes header and
context lines, there are "only" 1042 hunks.

Some would say that for a Committee consisting mostly of programmers,
programmers' representatives, and other kinds of geeks, that's a
surprisingly large number of hunks. Were there really 1024 of you?

Richard